JPH07121208A - Optimization design system - Google Patents

Abstract

PURPOSE:To enable a person who has neither sufficient experience nor knowledge about the optimization design system of a noise elimination and exhaustion system to securely utilize excellent experience and knowledge stored in the data base. CONSTITUTION:This system is equipped with an information data base 11 which previously stores constitution and performances to be referred to at the time of designing as to individual applicable components, a knowledge base 13 which stores a gathering of knowledge of optimum designing procedures, obtained experientially corresponding to the value ranges of the individual items included in the specifications of the design, according to a specific knowledge expressing method, a characteristic simulating means 15 which finds characteristics on the basis of simulation algorithm matching constitution designed on the basis of the optimum designing procedures as to the sound elimination and exhaustion system of the set constitution, and an inference means 17 which performs inference on the basis of knowledge to be applied by retrieving the knowledge according to specifications from the gathering of knowledge stored in the knowledge base 13, and refers to the information data base 11 and drives the characteristic simulating means 15 on the basis of the process procedure of the inference to design the constitution on the basis of the optimum designing procedures.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an optimization design system for optimally designing a noise reduction exhaust system of a vehicle in accordance with a desired required quality.

[0002]

2. Description of the Related Art In the process of developing and designing a muffler exhaust system for automobiles, the product characteristics given to consumers through safety, reliability, performance, operability and comfort related to operation, and appearance of the vehicle in which they are installed. In order to satisfy other required quality (or design specifications), optimization is performed from various viewpoints.
Furthermore, in such development design, a design method based on theoretical analysis has been partially introduced in recent years, but in fields where such a method cannot be applied, trial production and experiments are conducted in the same manner as in the past. Many thoughts and mistakes are being repeated.

Further, among the various items showing the above-mentioned required quality, in particular, the sound pressure loss and the sound volume of the exhaust system are
The external shape of the exhaust system is constrained by the underfloor shape of the vehicle, and it is technically contradictory because the model of the silencer and the mounting position must be selected within the range of the degree of freedom allowed under the constraint. Have a relationship. Furthermore, the phenomenon in which exhaust noise propagates while being attenuated non-linearly in an exhaust system actually mounted on a vehicle, or a phenomenon in which airflow noise is generated by pulsating flow has not yet been theoretically sufficiently clarified.

[0004] Therefore, the development design of the muffler exhaust system has sufficient experience on the above-mentioned technically contradictory relations and the design method based on theoretical analysis cannot be applied, and based on that experience, It was entrusted to a professional engineer who can make an accurate judgment.

[0005]

However, since the advanced and sufficient technical capabilities described above are cultivated based on many years of experience of engineers, it takes a lot of time and time to educate the personnel in charge of designing. Costly. Further, since each engineer does not always have sufficient experience and knowledge regarding both the muffling system and the exhaust system, a great restriction is imposed on the degree of freedom regarding the selection of the person in charge of design work.

[0006] It is an object of the present invention to provide an optimization design system capable of surely performing optimization design of a noise reduction exhaust system based on sufficient experience and knowledge even by a person who does not have such knowledge or experience.

[0007]

The present invention relates to an individual database applicable to a noise reduction exhaust system, an information database 11 for storing in advance a configuration and performance that can be referred to when designing the noise reduction exhaust system, and a design specification. Knowledge base 13 for accumulating a knowledge set of optimal design procedures obtained empirically corresponding to the range of each item included in
And a characteristic simulating means 15 for obtaining a characteristic based on a simulating algorithm adapted to the configuration of the noise elimination exhaust system having a configuration set based on the optimum design procedure, and a knowledge base 1.
The knowledge to be applied is searched based on the specifications from the knowledge set accumulated in 3, the information is inferred based on the knowledge, the information database 11 is referred to based on the inference processing procedure, and the characteristic simulation means 15 is driven. And an inference means 17 for performing a design based on the optimum design procedure by performing.

[0008]

In the optimum design system according to the present invention, the knowledge base 13 has a predetermined optimum design procedure consisting of sufficient experience regarding the design of the muffler exhaust system and knowledge including know-how found by the experience. It is systematically manifested and standardized and accumulated as a knowledge set based on the knowledge representation method. The inference means 17 retrieves knowledge to be applied based on the specifications from such a knowledge set, makes an inference based on the knowledge, and stores in advance in the information database 11 the parts to be adopted according to the processing procedure of the inference. The optimum design is performed by referring to the configuration and performance described above and obtaining the characteristic through the characteristic simulating means 15 for the noise reduction exhaust system having the configuration obtained by adopting the component.

That is, a muffler exhaust system having almost the same performance as a muffler exhaust system designed by an expert who has sufficient experience and knowledge described above can be reliably designed even for those who do not have such knowledge and experience. Therefore, it is not necessary to consider the abundance of experience and knowledge in the selection of designers while maintaining the performance. Therefore, the work efficiency can be improved, and the time and cost required for training the designers can be greatly reduced.

[0010]

Embodiments of the present invention will now be described in detail with reference to the drawings. FIG. 2 is a diagram showing an embodiment of the present invention.

In the figure, an inference engine 22, a simulation unit 23, a database generation unit 24, and a user interface unit 25 are connected to the knowledge base 21 via software and hardware interfaces, and the inference engine 22 and the user interface are further connected. 25 are also connected via software and hardware interfaces. Regarding software of each unit, the user interface unit 25 uses ART which is a tool for constructing a knowledge base, and the simulation unit 23 and the database generation unit 24 use FO.
Each is programmed using RTRAN77,
SYMBOLIC 3620, which is a LISP dedicated machine that operates under the operating system GENERAA, 3
630.

Regarding the correspondence between this embodiment and the block diagram shown in FIG. 1, the knowledge base 21 corresponds to the information database 11 and the knowledge base 13, the simulation unit 23 corresponds to the characteristic simulation means 15, and the knowledge Base 2
1, the inference engine 22 and the database generator 24 correspond to the inference means 17.

FIG. 3 is an operation flowchart of this embodiment. The operation of this embodiment will be described below with reference to FIGS. The knowledge base 21 stores the experience described above and the technology accumulated based on the experience (for example, the optimum design procedure regarding the basic performance design of the exhaust diameter, the defects already found and their countermeasures), along with the simulation unit 23. Systematically reveal and standardize the method of use, the method of analysis of simulation results obtained by using it, etc.
Knowledge of about 400 items represented in IF THEN (rule) format is stored in advance as a production rule base. Note that such a production rule base is a tool for building a knowledge base, ART and LIS.
Is programmed with the P function.

The knowledge stored in this way includes the demands of users and customers, the relationship between the standard quality of the developed product and the target cost, the product production method, the performance and reliability evaluation method, and the cost trial calculation method. , A muffler exhaust system development and design process, quality characteristics related to existing products, failure modes, and a wide range of other items. For example, for (1) silencer capacity (length),
Proposal for optimization (countermeasures) described below in the prescribed frequency band
If there is not, the current value is allowed and the standard value of the length of the next silencer is used as the increment, ..., (2) Regarding the basic structure of the silencer, the same optimization plan is applied in the prescribed frequency band. If it does not exist, it is discarded, and when adopting the next basic structure, the one with the smaller number of rooms is given priority .... (3) Regarding the generation of the silencing performance characteristic database described later, the amount of change in the variable part is the standard unit D or With unit length 1 (1 mm), changes in DS and IL due to structural changes in the silencer generate actual data for each pattern, frequency band, and site ...
(4) Regarding the difference in the optimum mounting position (boundary position) of the muffler, the insertion length of the inlet / outlet pipe obtained from the first best position must not be changed ... (5) The pipe diameter d and length l Regarding the change, when the diameter D is changed, the pressure loss is checked, and when the specification value is not satisfied, such a process involving the diameter change is not adopted, and the pipe diameter d and the length l are not changed. Change the value from standard size, D
Is performed in the above-mentioned standard unit D, l is performed in the above-mentioned unit length unit, ..., (6) Regarding the relationship between the development of tuning and the frequency band, when the required performance is satisfied in a predetermined frequency band, , Tuning is performed in a range lower than the frequency band, but the characteristics of the predetermined frequency band deteriorates by a predetermined value l ₁ or more, or the characteristics of a band adjacent to the frequency band deteriorates by a predetermined value l ₂ or more Is not adopted ...
(7) As the optimization ending condition, various knowledge is stored in advance, such as changing to the maximum permissible change amount and adopting the optimization plan that gives the best characteristics as a result.

Further, the knowledge base 22 includes the shapes (including the cross-sectional shape of the pipe material, the dimensions of the plate-like body and the plate thickness of the pipe material, etc.), standards, and engine types of standard parts to be used in the muffler exhaust system. (Including the number of cylinders, number of cycles, displacement, etc.), basic pattern of internal structure of the silencer that can be adopted (numerical code indicating the internal shape and internal exhaust gas flow for each already produced silencer) Is stored in advance as an information database. Further, in the information database, a value that can be designated by a character string as a required performance (design specification) item by a designer is stored in advance as a keyword indicating the value with short characters. In addition, such an information database is
It is programmed using RT.

The operation described below transits in chronological order and is described later in "Required performance input phase", "Silent silencer internal structure basic pattern selection phase", "Silent performance characteristic database generation phase" and "Optimization". It consists of four phases called "Phase". Regarding the processing procedure (algorithm) in each of these phases, the knowledge stored in advance in the knowledge base 21 as the production rule base is used as the inference engine 22.
Are determined by appropriately reading and analyzing, and are realized through the knowledge base 21. Therefore, in the following,
The operation of the inference engine 22 is omitted for simplicity.

Further, the knowledge base 21 has a communication control program programmed by LISP, and performs interactive communication by facing the simulation unit 23 and the database generation unit 24 through the program, and detailed operation thereof. The detailed description is omitted here for simplicity.

[Required Performance Input Phase] The user interface unit 25 has an operation / display unit (not shown).
A menu configured by arranging icons (see, for example, FIG.
The list of parts shown in (1) is displayed, the required performance specified by the designer in an interactive format is fetched, and the individual items included in the specification are given to the knowledge base 21. Items specified as required performance include discharge noise performance, pressure loss performance, external shape of the exhaust system, etc., in accordance with the class of the vehicle in which the noise reduction exhaust system is to be installed and the type of engine to which it is applied. included. The knowledge base 21 determines whether or not each of the items given in this manner is appropriate with respect to the mutual consistency with other items and the conformity with the standard, and the knowledge base 21 communicates with the designer through the user interface 25. Have an intellectual dialogue between.

Further, the user interface section 25 is
When the designer specifies a keyword as the item value of the required performance, the keyword is sequentially converted to the corresponding set value via the knowledge base 21, and the above-described determination and other processing are performed. Therefore, the designer is provided with an operating environment in which the required performance can be efficiently specified while preventing erroneous input by a simple operation of clicking an icon or inputting a keyword.

[Phase of selecting basic pattern of silencer internal structure] When all items of the required performance are input in this way, the knowledge base 21 subdivides each item within a range that can satisfy the required performance. By performing the correction, the target value to be adopted in the process of optimum design described later is determined (FIG. 3).

Next, the knowledge base 21 gives the target value thus determined to the simulation section 23, and the simulation section 23 determines the optimum values of the mounting position of the silencer and the internal boundary conditions based on the target value. Ask. The knowledge base 21 takes in the optimum value of the boundary condition obtained in this way, and refers to the internal structure basic pattern of the silencer stored as the above-mentioned information database based on the boundary condition to assist the adoption. A candidate for the basic pattern of the internal structure of the silencer (primary muffler) is obtained. If there are multiple such candidates, the knowledge base 2
In No. 1, the error of the muffling characteristic of each candidate with respect to the desired muffling characteristic is weighted by the degree of importance for each predetermined frequency band, arithmetically comprehensively compared, and the candidates are narrowed down based on the result of the comparison.

Further, the knowledge base 21 fixes the candidate of the auxiliary silencer thus obtained, and determines the number of rooms in the internal structure of the main silencer (main muffler) to be fixed and combined. The same weighting is performed while giving priority to the smaller one, and comprehensive comparison is performed, and the internal boundary condition is optimized based on the result of the comparison. Hereinafter, the knowledge base 21 repeats the same optimization process until the convergence condition that satisfies the target value is reached, and selects the internal structure basic pattern of the silencer (FIG. 3).

[Muffling Performance Characteristic Database Generation Phase] When the basic pattern of the internal structure of the muffler is selected in this way, the knowledge base 21 is equipped with an exhaust system equipped with the muffler having the selected basic pattern. For each variable part that can be changed with (for example, the diameter of the exhaust system tube, the diameter and length of the pipe installed in the silencer body, room allocation in the silencer, etc.) However, the simulation unit 23 is requested to calculate the silencing characteristics and the pressure loss of each frequency band for each variation part and variation amount. The amount of fluctuation is set using a standard value in consideration of standardization.

The simulation unit 23 responds to such a request by a predetermined calculation method (for example, SAE (Society o
f Automotive Engineers) No. 890619. ) Is applied, the insertion loss, the transmission loss, the discharge noise level, the optimum insertion length of the pipe installed at the inlet and outlet of the muffler, and the pressure loss characteristic are sequentially calculated, and based on the calculation results, Determine the muffling characteristics and pressure loss. Further, in the process of the calculation of such a numerical simulation, the simulation unit 23 determines, for example, as shown in FIG. 5, the structure and the specification value of the silencer, which is the target of the simulation, in the knowledge base 21.
The progress status of the work is notified to the operator by sequentially displaying it via the user interface unit 25.

The knowledge base 21 gives the silencing characteristics and pressure loss thus obtained to the database generating section 24, and the database generating section 24 collects these silencing characteristics and pressure loss to generate a silencing performance characteristic database. (Fig. 3).

[Optimization Phase] The knowledge base 21 confirms the exhaust system equipped with the silencer having the basic structure selected in the selection phase of the internal structure basic pattern of the silencer, and attaches the exhaust system to a desired engine. Numerical information indicating the closed state is given to the simulation unit 23. The simulation unit 23 obtains the prediction performance according to the numerical information (see FIG. 6).
), The knowledge base 21 determines whether or not the predicted performance satisfies the required performance (FIG. 6). Here, the main items of the predictive performance consist of the level of discharge noise and the pressure loss characteristics, but the simulation unit 23 obtains higher-order frequency components of the discharge noise in parallel when obtaining such predictive performance. And give it to the knowledge base 21.

When the knowledge base 21 recognizes that the required performance is satisfied by such a determination, it completes the optimization phase, but on the contrary, when it is not satisfied, the above-mentioned higher-order frequency component is added. Based on the discharge noise that is a factor that the required performance is not satisfied based on the peak level,
Calculate the frequency, the amount of attenuation required to satisfy the required performance, the engine speed corresponding to the maximum allowable level, etc.,
Further, it is determined whether or not each of these items is a problem and the problem is extracted (FIG. 6).

Next, the knowledge base 21 corrects the target values of the insertion loss characteristic and the pressure loss performance of the silencer based on each item of the problems extracted in this way (FIG. 6).
Further, the knowledge base 21 obtains the silencing characteristic and the pressure loss characteristic corresponding to the variation amount of the variable portion by referring to the silencing performance characteristic database generated in advance (FIG. 6), and based on these characteristics. Adjust the internal structure of the silencer and the diameter of the exhaust system tube (Fig. 6).

The knowledge base 21 gives a priority order to each optimization plan obtained by such an adjustment, and provides the simulation section 23 with numerical information indicating the optimization plan corresponding to the priority order. The simulation unit 23 obtains the prediction performance according to the numerical information, and the knowledge base 21 sequentially determines whether the prediction performance satisfies the required performance, and the same is performed until it recognizes that the required performance is achieved. The above process is repeated (FIG. 6). Further, when confirming that the required performance is achieved in this way, the knowledge base 21 comprehensively confirms the required performance for the bands other than the frequency band in which the above-mentioned problem has occurred through the simulation unit 23. The optimization processing result obtained by the confirmation is output through the user interface unit 25 to complete the optimization phase (FIG. 3).

The user interface unit 25, as shown in FIG. 7, based on the processing results (internal structure of the muffler, diameter of the exhaust pipe, etc.) thus provided, the structure of the muffler necessary for drawing work. Of the muffler that gives the basic pattern, and the like. Further, as shown in FIG. 8, regarding the discharge sound level, the target value given by the required specifications and the value obtained based on the simulation are displayed. It is displayed according to the engine speed.

Further, the sound deadening exhaust system optimally designed through such a series of processes has a displacement of 2000 cc, for example.
When applied to a 6-cylinder, 4-cycle normally aspirated engine, the discharge noise level is about 5 dB compared to the design example by a conventional expert, as shown in Fig. 9 (a).
Although it is high, the change of the discharge noise level with respect to the engine speed matches the design example of the expert with accuracy,
As shown in FIG. 9 (b), the technically contradictory pressure loss with respect to the volume of silence was about 11%.

As described above, according to the present embodiment, the experience of the expert regarding the design of the exhaust system and the knowledge (including know-how) cultivated by the experience are systematically realized and standardized to form the knowledge base. In addition, we utilize the results of numerical simulation of acoustic performance as a criterion for inference based on the knowledge base.

In other words, even a person who does not have such experience or knowledge can reliably design a muffler exhaust system having substantially the same performance as a design example based on the experience or knowledge, so that the person in charge of designing has experience. It is not necessary to take into consideration the abundance of knowledge and knowledge, the work efficiency can be improved, and the time and cost required for the personnel training of the designer can be greatly reduced.

In this embodiment, after selecting the basic pattern of the internal structure of the muffler, the characteristics corresponding to each variable part that can be varied by the exhaust system equipped with the muffler of the selected basic pattern are set. Although it is collectively obtained and generated as the silencing characteristic performance database, the present invention is not limited to such a method, and for example, the characteristic may be obtained by appropriately driving the simulation unit 23 in the optimization phase.

Further, in the present embodiment, the knowledge accumulated in the knowledge base 21 is represented based on the production rules, but the present invention is not limited to such production rules, and an adapted inference engine is installed. If possible, any knowledge representation method such as a semantic network, a frame, an axiomatic set of predicate logic, or the like may be adopted.

Further, in this embodiment, the inference engine 22 is used.
Makes an inference by sequentially searching and translating the knowledge accumulated in the knowledge base 21, but the present invention is not limited to such a configuration, and for example, knowledge described by a production rule is compiled in advance. You may try to speed up reasoning.

Further, in this embodiment, nothing is described about updating the knowledge accumulated in the knowledge base 21 and the like, but for example, the user interface unit 25 is provided with interface means exclusively for development or the silencer. Means for maintaining an information database relating to the basic pattern of the internal structure may be provided, or new knowledge accumulated during the operation of the optimization design system according to the present embodiment may be appropriately added for training.

[0038]

As described above, according to the present invention, those who have sufficient experience and knowledge can evaluate and optimize the characteristics based on the simulation algorithm adapted to the structure and performance of the adopted parts. However, it is possible to reliably design a muffling exhaust system having the same performance as an exhaust muffling system designed by an expert based on these experiences and knowledge. Also, as long as it is operated by a person who has basic knowledge about the muffler exhaust system,
Conventionally, quality deterioration and work rework that may occur due to mistakes of experts during designing are greatly reduced.

Therefore, the man-hours required for the optimized design, trial manufacture and experiment of the noise reduction exhaust system are greatly reduced while the design quality is made uniform and improved, and the design development period is shortened and the working environment involved in the design work is improved. It comes off.

[Brief description of drawings]

FIG. 1 is a principle block diagram of the present invention.

FIG. 2 is a diagram showing an embodiment of the present invention.

FIG. 3 is an operation flowchart of this embodiment.

FIG. 4 is a diagram showing a method of inputting development specifications.

FIG. 5 is a diagram showing an example of information displayed during a simulation.

FIG. 6 is a flowchart of optimization processing.

FIG. 7 is a diagram (1) showing a design example of a noise reduction exhaust system obtained by applying the present embodiment.

FIG. 8 is a diagram (2) showing a design example of the noise reduction exhaust system obtained by applying the present embodiment.

FIG. 9 is a diagram showing evaluation results of design examples obtained by applying the present embodiment.

[Explanation of symbols]

 11 information database 13 knowledge base 15 characteristic simulation means 17 inference means 21 knowledge base 22 inference engine 23 simulation section 24 database generation section 25 user interface section

Claims (1)

[Claims] 1. An information database (11) for storing in advance a configuration and performance that can be referred to when designing a noise reduction exhaust system for each part applicable to the noise reduction exhaust system, and an individual data base included in the design specifications. A knowledge base (13) for accumulating a knowledge set of optimal design procedures obtained empirically corresponding to the range of items based on a predetermined knowledge representation method, and a silencing exhaust having a configuration set based on the optimal design procedures. With respect to the system, characteristic simulation means (15) for obtaining characteristics based on a simulation algorithm adapted to the configuration, and knowledge to be applied based on the specifications are searched from the knowledge set accumulated in the knowledge base (13). The optimum design procedure is performed by inferring based on the knowledge, referring to the information database (11) and driving the characteristic simulating means (15) based on the processing procedure of the inference. Optimization design system comprising the inference means (17) for performing design based.